|
|
Boost Testing : |
From: Caleb Epstein (caleb.epstein_at_[hidden])
Date: 2006-11-08 15:15:50
On 11/8/06, Rene Rivera <grafikrobot_at_[hidden]> wrote:
> Caleb Epstein wrote:
>
> > Very puzzling. When I run "bjam --v2 -d2" from the
>
> Most likely it means that g++ returned an error value without saying
> anything.
I've never seen that happen, and I don't think its happening here.
> > If I take this exact same "/openpkg/bin/g++" command and paste it at
> > my shell prompt, it succeeds. Very weird!
>
> Are you sure? What is the exit status value of the command?
When I run it by hand, I get an exit status of 0 and the compiler
generates the object file.
I ran the bjam command under "truss" (system call tracer) and it
appears that the gcc stages all exit with status 0 even though bjam
says an error occurred:
Some excerpts from the truss logfile:
bjam starts g++:
12234: fork1() = 12236
12236: fork1() (returning as child ...) = 12234
12236: execve("/openpkg/bin/g++", 0x08077630, 0x08077668) argc = 13
<lots of stuff, but ultimately>
12236: _exit(0)
12234: waitid(P_PID, 12236, 0x08047840, WEXITED|WTRAPPED|WNOWAIT) = 0
12234: waitid(P_PID, 12236, 0x08047840, WEXITED|WTRAPPED) = 0
12234: _exit(0)
12215: waitid(P_ALL, 0, 0x080469D0, WEXITED|WTRAPPED) = 0
12215: write(1, " . . . f a i l e d g c".., 122) = 122
12215: write(1, " . . . r e m o v i n g ".., 105) = 105
12215: write(1, " . . . f a i l e d u p".., 31) = 31
So all of the sub-processes appear to exit with status 0, but bjam
insists an error occurred.
-- Caleb Epstein